Product video ads with Veo 3.1: workflow and real cost

Plan the ad shot by shot

A common structure for a short product ad is four beats. Each generated beat is one clip; the last one is better made by hand.

The Lite vs Fast vs Quality comparison explains what the dearer models add. For product work, starting from a real photo usually matters more than the model, and the Veo 3.1 image-to-video guide covers first and last frames in detail.

What an ad costs

The end card costs nothing because you make it in your editor. Budget for retakes on the reveal shot, where label accuracy matters most: if it takes two attempts, its cost doubles. Failed generations are refunded automatically, but a finished clip you don’t like is not a failure. With sound only on the hook, the 30-second ad above costs 28 credits. The cost calculator prices any other mix.

Workflow: product photo to finished ad

1. 1

   Shoot or pick the product photos

   One clean packshot for the reveal and, if you have it, one lifestyle photo for the in-use beat. Plain background, even light, label facing the camera. PNG or JPEG, up to 20 MB each.
2. 2

   Upload it as the First Frame Image

   In the generator (sign in with Google; no card is needed to sign up), switch to the Image-to-Video tab and upload the photo. Add your packshot as the Last Frame Image if the shot should end on it.
3. 3

   Describe the motion, not the product

   The photo already shows what the product looks like. Write one action and one camera move, the lighting, and “no text, no logos”, for example “slow push-in, the bottle turns a quarter turn, soft window light”.
4. 4

   Set the controls

   Model: Veo 3.1 Fast · Resolution: 1080p · Aspect Ratio: 9:16 · Duration: 8s · With Audio: off. That is 6 credits per clip; the cost shows next to Generate Video before you start.
5. 5

   Check the label, then assemble

   Scrub through the clip and check the label and packaging in every frame; regenerate with simpler motion if they warp. Then join the beats, add music, captions and your end card in your editor.

A real ad series from our gallery

One user shared several scenes of a vertical snack ad, each written as a JSON shot list with one voiceover sentence. Three of them show how a listicle-style ad breaks into clips, including one beat (a text-only card) that is better made in an editor. All three files have a sound track.

• UGC snack scene with a voiceover (JSON)

  Veo 3.1 Lite · 9:16 · 720p · 4 s · audio on · 2 credits (about $0.43–$0.50)

  This uses the same JSON template as the pouch-reveal scene. “dialogue_voiceover” holds the line and “sound_design” asks for continuous music, so every scene in the series is built the same way.

• Snack ad, scene 8: a text-only “#3” card (JSON)

  Veo 3.1 Lite · 9:16 · 720p · 4 s · audio on · 2 credits (about $0.43–$0.50)

  It shows how the series planned a listicle ad: numbered point cards between product shots, each with one voiceover sentence.

• Snack ad, scene 9: on-the-go shot with a voiceover line (JSON)

  Veo 3.1 Lite · 9:16 · 720p · 4 s · audio on · 2 credits (about $0.43–$0.50)

  Scene 9 of the same snack series as the voiceover and reveal scenes: two quick shots (walking, then a pouch raised to camera) and one short voiceover line, written in the same JSON fields as the other scenes so the series stays consistent.

More product shots, and one to learn from

These are text-to-video clips shared by users, with settings measured from each file. The mug prompt is here as a warning: it asks the model to copy a product it never saw.

• Colour-changing mug “from a product reference” (text only)

  Veo 3.1 Fast · 16:9 · 1080p · 8 s · audio on · 10 credits (about $2.13–$2.50)

  It is a well-organised product brief: subject, action, setting, light, camera and style each get their own line, and a closing list of strict requirements rules out overlays and logos.

• A handful of miniature pottery poured onto linen

  Veo 3.1 Fast · 9:16 · 1080p · 8 s · audio on · 10 credits (about $2.13–$2.50)

  A tactile texture shot: one hand, one pour, one cut to a second drop, on a plain white surface with soft light that shows off a glossy finish. Small handmade goods often sell on exactly this kind of shot.

• The wet-boots idea in 4 seconds at 720p

  Veo 3.1 Lite · 9:16 · 720p · 4 s · audio off · 1 credit (about $0.21–$0.25)

  This is a shorter take on the same idea: one action (drop the boots, squeeze the tongue) fits a 4-second clip. Lite at 720p is the cheapest way to find out whether an idea works before you pay for a longer, sharper take.

• Problem shot: back pain at a desk

  Veo 3.1 Lite · 16:9 · 1080p · 8 s · audio on · 5 credits (about $1.07–$1.25)

  One sentence covers who it is, what’s wrong, the camera move and the lighting. That is enough for a “before” shot in an ad for a chair, a cushion or a physio service.

• One spoken line, no scene description (Spanish)

  Veo 3.1 Fast · 16:9 · 1080p · 8 s · audio on · 10 credits (about $2.13–$2.50)

  This shows what dialogue alone gets you: the model has to invent the speaker, the setting and the shot. It’s a quick way to hear a line read, but for an ad you would add who says it and where.

For cinematic packshots and texture shots with full prompts and copy buttons, see the cinematic and product shot prompts.

Limits for product ads

The limits every workflow shares (8-second clips, 16:9 and 9:16 only, unreliable generated text, drift between clips, optional sound, rights) are listed once on the use cases page. These are the ones specific to this kind of video:

• Text-only prompts invent the product. Several gallery prompts refer to an uploaded image or a “provided product reference” but were run as text-to-video, so the model never received it; the mug prompt above is one. Use Image-to-Video when the product must match.
• Labels can warp. Even from a photo, small text on packaging can change as the camera moves. Check every frame, and if it changes, retry with slower, simpler motion.
• No fake customers. A generated person holding your product is fine as a dramatisation. Captioning them as a real customer or review is a fake testimonial.
• Uploads stay private. Videos made from your uploaded images are never shared to the public gallery.

Frequently asked questions